Disaster Recovery Log-Based Data Replication Setup
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing disaster recovery systems face delays in starting log application at remote sites due to long initial copy times, leading to data inconsistency and log wrap round issues when online processing is delayed, causing substantial delays in data replication.
Innovation Solution
A disaster recovery system that allows initial copy to be executed without stopping online processing by using a dual-volume setup where data is copied sequentially and prioritized, ensuring that logs are applied only after data is fully copied, thereby maintaining data consistency and preventing log wrap round.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If initial copy is executed by stopping online processing, then data consistency is ensured, but online processing is delayed
Solution Approach 1:
The storage device is divided into a first storage area for online processing and a second storage area for initial copy. This segmentation allows online processing to continue in the first area while initial copy is performed in the second area, resolving the contradiction between data consistency and processing continuity.
Solution Approach 2:
The system performs preliminary actions by copying data to the second storage area before stopping online processing. This preliminary copy ensures that when online processing must be stopped for maintenance or migration, data consistency can be maintained without losing productivity during the transition period.
2Quantity of substance
If initial copy takes a long time, then data replication is complete, but log application start is delayed causing data inconsistency
Solution Approach 1:
The storage device is segmented into areas for initial copy and log application. The log applying unit operates independently on the second storage area after initial copy, allowing parallel processing and eliminating the delay where log application would wait for initial copy completion.
Solution Approach 2:
The system maintains continuous useful action by allowing log application to proceed independently after initial copy. The log applying unit continuously applies logs to the second storage area without interruption, ensuring data consistency is maintained while avoiding time loss from sequential processing.
3Manufacturing precision
If storage system is put in quiescent state for initial copy, then data copying is accurate, but online processing must be put off
Solution Approach 1:
The storage device is divided into a first storage area that remains active for online processing and a second storage area that is quiescent for initial copy. This segmentation allows accurate data copying in the second area while maintaining processing availability in the first area, resolving the contradiction between copying precision and productivity.
Data Source
AI summary
Provided is a technique of setting up a disaster recovery system. In a log-based disaster recovery system including a first site for processing data and a second site for storing a backup of the first site, the first site has a volume where data is stored and a volume that is made quiescent. The disaster recovery system accepts data updates while sequentially copying initial data from the quiescent volume to the second site. A copy of a log of the first site is kept in the second site and is synchronized with the original log. When data of the first site is updated and corresponding initial data has not been copied to the second site, the data to be updated is preferentially copied to the second site. A log is applied to the copied data, thereby matching data of the second site with data of the first site.